A factory environment monitoring system and method integrating a sensor network

Abstract

This invention relates to the field of factory management technology, specifically to a factory environmental monitoring system method integrating sensor networks. The method includes: collecting multi-dimensional environmental data within the factory using sensors and designing data acquisition strategies for event-triggered sampling and adaptive sampling; constructing a sensor self-organizing network protocol to achieve automatic sensor network discovery and dynamic networking; creating a link quality adaptive mechanism that automatically adjusts transmission power or switches channels when the link deteriorates by periodically detecting link quality indicators; extracting features from the collected environmental data and constructing a machine learning model to predict environmental change trends; and constructing an environmental regulation strategy to pre-adjust the environment based on predicted trend anomalies. This invention enables the integrated intelligent sensing, prediction, and control of the factory environment, improving environmental management and ensuring production safety and efficiency.

Description

Technical Field

[0001] This invention relates to the field of factory management technology, specifically to a factory environmental monitoring system and method integrating a sensor network. Background Technology

[0002] Most existing factory environmental monitoring systems use wired sensors, which are costly to deploy, inflexible, and difficult to achieve large-scale, multi-dimensional environmental monitoring. Furthermore, traditional data acquisition strategies use fixed time intervals, failing to consider the dynamic characteristics of environmental changes, resulting in low data acquisition efficiency.

[0003] Furthermore, existing sensor network deployment methods typically employ static configurations, which cannot adapt to changes in network topology such as the addition or removal of nodes, resulting in poor network scalability. During wireless transmission, the complex and variable industrial environment makes wireless link quality susceptible to interference, leading to poor data transmission reliability.

[0004] In terms of data processing and analysis, existing methods mainly rely on manual statistics and threshold comparison, which lack real-time performance and intelligence. They are unable to detect environmental change trends in a timely manner and lack the ability to provide early warnings and proactive adjustments for abnormal situations, which may lead to production safety and efficiency problems.

[0005] Existing factory environmental monitoring systems suffer from poor network adaptability and lagging data analysis, necessitating a new, low-cost, flexible, efficient, and intelligent environmental monitoring solution.

[0006] In view of this, the present invention proposes a factory environment monitoring system and method integrating sensor networks. Summary of the Invention

[0007] To achieve the above objectives, the present invention provides a factory environmental monitoring system and method integrating a sensor network, the specific technical solution of which is as follows:

[0008] A factory environment monitoring method integrating sensor networks, comprising:

[0009] Multi-dimensional data of the factory environment is collected through sensors, and an adaptive data acquisition strategy is constructed, including an event-triggered sampling strategy and an adaptive sampling strategy.

[0010] Build a sensor self-organizing network strategy to automatically discover sensor networks and dynamically form networks;

[0011] Create a link quality adaptive mechanism to periodically detect wireless link signal indicators. When the link quality drops below a threshold, adaptively adjust the transmit power or switch channels.

[0012] Extract features from environmental data collected by sensors, build and train a machine learning model to predict the trend of environmental data change. Extract features from real-time collected environmental data and use the trained machine learning model to predict the trend of environmental data change.

[0013] Develop a strategy to adjust the factory environment in advance when the environmental data predicted by the machine learning model shows abnormal trends.

[0014] Preferably, the sensors used in the factory include temperature sensors, humidity sensors, infrared optical sensors, acoustic sensors, and gas sensors.

[0015] An adaptive data acquisition strategy is constructed, comprising an event-triggered sampling strategy and an adaptive sampling strategy; the adjustment step size of the sensor sampling time interval is defined as t0, and the initial sampling interval is t; the current sampling interval is set to t. i The current data is d i The previous data was d i-1 ;

[0016] The event-triggered sampling strategy includes: when When the sensor is triggered to sample, the sampled data is updated; at the same time, the current sampling interval t is changed. i The value is reduced by t0, but not less than the set minimum sampling interval t. min ;where δ (g) Let g be the preset threshold for the g-th type of sensor, where g ∈ [1, G] and G is the total number of sensor types.

[0017] The adaptive sampling strategy includes: when m consecutive samples satisfy... At that time, the current sampling interval t will be... i The value increases by t0, but does not exceed the set maximum sampling interval t. max ;

[0018] consecutive m samples satisfy At that time, the current sampling interval t will be... i The value is reduced by t0, but not less than the set minimum sampling interval t. min Where m is a preset sensitivity parameter.

[0019] Preferably, a self-organizing network protocol is constructed, in which each sensor is regarded as a network node, and the network node includes a parent node network composed of network coordinators and a child node network composed of sensors;

[0020] Dynamic addresses are assigned to sensors. The network coordinator presets a network address pool, calculates the network address assigned to each sensor, and performs network address conflict detection on the sensors.

[0021] The system enables automatic sensor discovery and dynamic networking. After a new sensor is powered on, it first enters a network listening state and scans multiple channels for a predetermined time. If no network beacon frames are received, the new sensor acts as a network coordinator, creates a new network on an idle channel, and periodically broadcasts beacon frames. If the new sensor receives a network beacon frame, it parses the network parameters in the beacon, joins an existing network, and requests address allocation.

[0022] Preferably, a link quality assessment metric is defined, and a signal strength indicator (RSSI) is defined. a , represents the signal strength received by sensor a; the signal-to-noise ratio (SNR) is defined. a , represents the signal-to-noise ratio received by sensor a; defines the packet loss rate. This represents the packet loss rate from sensor a to network coordinator a0;

[0023] Link quality is periodically monitored. Each network coordinator periodically broadcasts link probe packets, which include the node number, transmit power, and timestamp. Upon receiving the link probe packets, the child node sensors of the network coordinator measure the Signal Strength Indicator (RSSI) and Signal-to-Noise Ratio (SNR), and send the measurement results, along with their own node number and timestamp, to the parent network coordinator. After receiving the response from child node sensor a, the parent network coordinator a0 calculates the packet loss rate. Where, N sent N represents the total number of packets sent from parent node network coordinator a0 to child node sensor a. received This represents the number of packets successfully received by child node sensor a.

[0024] Define link quality thresholds, including: Signal Strength Indicator (RSSI) threshold. th Signal-to-noise ratio (SNR) threshold th and packet loss rate threshold PLR th When the signal strength indicator RSSI a <RSSI th When the link signal strength is insufficient, the signal-to-noise ratio (SNR) is also insufficient. a <SNR th When the link signal-to-noise ratio is insufficient, the packet loss rate will be high. If this happens, the packet loss rate of the link will be too high.

[0025] Preferably, the transmit power between the sensor and the network coordinator is adaptively adjusted. When the parent node network coordinator a0 detects that the link quality with the child node sensor a does not meet the quality threshold condition, the transmit power is increased.

[0026] Define the transmit power adjustment step size ΔP, and let the initial transmit power be P0; when RSSI is satisfieda <RSSI th or SNR a <SNR th If the link quality still does not meet the quality threshold condition after increasing the transmit power, continue to increase it until the maximum transmit power P is reached. max ;

[0027] When child node sensor a detects that the link quality with parent node network coordinator a0 does not meet the quality threshold condition and has reached the maximum transmit power, the channel is switched.

[0028] Define the candidate channel set C = {c1, c2, ..., c K}, where K is the total number of candidate channels; the child node sensor a randomly selects a candidate channel c from the candidate channels. k Switch to the alternative channel for communication;

[0029] If the link quality still does not meet the quality threshold condition after switching channels, random switching will continue until all alternative channels have been switched.

[0030] If all alternative channels fail to meet the quality threshold conditions, then the child node sensor a sends a network reorganization or node reassociation request to the parent node network coordinator a0.

[0031] A link quality recovery mechanism is constructed. When the child node sensor a detects that the link quality with the parent node network coordinator a0 has recovered to above the quality threshold, the transmit power is reduced successively, with each reduction of transmit power being ΔP, until the initial transmit power P0 is reached. If the link quality deteriorates again after reducing the transmit power, the adaptive adjustment of transmit power or channel switching procedure is re-executed.

[0032] Preferably, historical environmental data is acquired, and the characteristics of the environmental data collected by the sensor are defined, including: temperature data characteristics, humidity data characteristics, infrared optical data characteristics, acoustic data characteristics, and gas concentration data characteristics.

[0033] The temperature data features include: average temperature Temperature variance and the temperature change rate T'; the humidity data characteristics include: average humidity Humidity variance and humidity change rate H'; the infrared optical data features include: average light intensity Light intensity variance And the rate of change of light intensity I'; the acoustic data characteristics include: average sound pressure level sound pressure level variance and the sound pressure level change rate P'; the gas concentration data characteristics include: average gas concentration gas concentration variance And the rate of change of gas concentration C';

[0034] The sliding window method is used to extract time-series features from historical environmental data, a machine learning model is constructed, and a Long Short-Term Memory (LSTM) network model is used to model and predict environmental data.

[0035] Preferably, the input to the LSTM model is the extracted environmental data time series features X = (x1, x2, ..., x...). E ), where E is the number of time windows in the historical environmental feature dataset; the output of the LSTM model is the predicted environmental data for the next F time windows. in Let f be the predicted environmental data value for the f-th time window, where f∈[1,F];

[0036] The acquired historical environmental feature data is divided into a training set and a test set. The LSTM model is trained using the training set and evaluated using the test set. During training, mean squared error is used as the loss function.

[0037] The trained LSTM model is used to predict real-time acquired environmental data. The specific steps are as follows: preprocessing and feature extraction are performed on the real-time acquired environmental data to obtain the current time series windowed feature vector x. t , will x t The data is input into an LSTM model to obtain predicted environmental data for the next F time windows. Predicted values The data is compared with a set threshold for abnormal trends in environmental data to determine whether the trend of environmental data changes is abnormal.

[0038] Preferably, anomaly trend thresholds for environmental data change trends are defined, including a temperature change trend threshold ΔT. th Humidity change trend threshold ΔH th Infrared light intensity change trend threshold ΔI th Sound pressure level change trend threshold ΔP th and the threshold value ΔC for the trend of gas concentration change th ;

[0039] The trained LSTM model is used to predict environmental data for the next F time windows, resulting in a sequence of predicted values. in These represent the predicted values ​​for temperature, humidity, infrared light intensity, sound pressure level, and gas concentration for the f-th time window, respectively.

[0040] Preferably, the trend of environmental data changes within a time window includes: temperature change trend. Humidity change trend Infrared light intensity variation trend Sound pressure level change trend Gas concentration change trend

[0041] Perform anomaly detection on the predicted trend sequence: If The predicted temperature change trend is abnormal; if The predicted humidity change trend is abnormal; if The predicted trend of infrared light intensity variation is abnormal; if The predicted sound pressure level change trend is abnormal; if The predicted trend of gas concentration change is abnormal;

[0042] If any outlier appears in the predicted trend sequence, the factory environment change trend adjustment strategy will be implemented to restore the corresponding environmental data change trend to within the outlier trend threshold.

[0043] After environmental adjustments, the LSTM model continues to be used to predict future environmental data. If the predicted trend remains abnormal and alarm conditions are met, an alarm is issued. The alarm conditions are as follows:

[0044]

[0045] Where I(·) is the indicator function; This represents the predicted trend of environmental data changes in the f-th time window, where X represents any type of environmental data, and ΔX th α is the corresponding abnormal trend threshold; α is the abnormal proportion threshold.

[0046] A factory environment monitoring system integrating a sensor network, used in the aforementioned factory environment monitoring method integrating a sensor network, includes: a data acquisition strategy module, an automatic networking module, a network signal conditioning module, an environmental data prediction module, and an environmental conditioning module;

[0047] The data acquisition strategy module collects multi-dimensional data of the factory environment through sensors and constructs an adaptive data acquisition strategy, including an event-triggered sampling strategy and an adaptive sampling strategy.

[0048] The automatic networking module is used to construct a sensor self-organizing network strategy, adaptively discover sensor networks, and dynamically form networks.

[0049] The network signal conditioning module is used to create a link quality adaptive mechanism, periodically detect wireless link signal indicators, and automatically adjust the transmission power or switch channels when the link quality drops below a threshold.

[0050] The environmental data prediction module is used to extract features of environmental data collected by sensors, construct and train a machine learning model to predict the trend of environmental data changes. It extracts features from real-time collected environmental data and uses the trained machine learning model to predict the trend of environmental data changes.

[0051] The environmental adjustment module is used to construct a factory environment change trend adjustment strategy. When the environmental data change trend predicted by the machine learning model is abnormal, the factory environment is pre-adjusted.

[0052] The beneficial effects of this invention are: by comprehensively collecting factory environmental data, this invention can efficiently obtain key information, optimize data collection strategies, reduce energy consumption, and provide data support for environmental monitoring.

[0053] This invention simplifies deployment and maintenance, improves network flexibility and scalability, and optimizes sensor network performance by automatically building sensor networks.

[0054] This invention ensures data transmission reliability, enhances the robustness of sensor networks, and guarantees real-time reporting of critical data by adaptively adjusting link quality.

[0055] This invention predicts environmental change trends and detects anomalies in advance, providing a basis for environmental regulation decisions and improving the foresight and proactivity of environmental management.

[0056] This invention automatically adjusts environmental parameters based on predictions, reduces the impact of anomalies on production, improves environmental management, and ensures production safety and efficiency. Attached Figure Description

[0057] Figure 1 A flowchart of a factory environment monitoring method integrating a sensor network provided by the present invention;

[0058] Figure 2 This invention provides a structural diagram of a factory environmental monitoring system with an integrated sensor network. Detailed Implementation

[0059] To better understand the invention, various aspects of the invention will be described in more detail with reference to the accompanying drawings. It should be understood that these detailed descriptions are merely illustrative of exemplary embodiments of the invention and are not intended to limit the scope of the invention in any way. Throughout the specification, the same reference numerals refer to the same elements. The expression "and / or" includes any and all combinations of one or more of the associated listed items.

[0060] It should also be understood that expressions such as "comprising," "including," "having," "containing," and / or "comprising" are open-ended rather than closed-ended expressions in this specification, indicating the presence of the stated features, elements, and / or components, but not excluding the presence of one or more other features, elements, components, and / or combinations thereof. Furthermore, when expressions such as "at least one of..." appear after a list of listed features, they modify the entire list of features, not just individual elements in the list. Additionally, when describing embodiments of the invention, the word "may" is used to mean "one or more embodiments of the invention." And the term "exemplary" is intended to refer to examples or illustrations.

[0061] Unless otherwise specified, all terms used herein (including engineering and technical terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that, unless expressly stated herein, terms defined in common dictionaries shall be interpreted as having the meaning consistent with their meaning in the context of the relevant art, and not in an idealized or overly formalized sense.

[0062] It should be noted that, unless otherwise specified, the embodiments and features described in this invention can be combined with each other. The invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0063] Example 1

[0064] Reference Figure 1 This is the first embodiment of the present invention, which provides a factory environment monitoring method integrating a sensor network.

[0065] S1: Collect multi-dimensional data of the factory environment through sensors and build an adaptive data acquisition strategy, including an event-triggered sampling strategy and an adaptive sampling strategy.

[0066] In the factory, sensors, including temperature sensors, humidity sensors, infrared optical sensors, acoustic sensors, and gas sensors, are used. Each type of sensor is deployed in the corresponding location as required, and the sensors transmit the collected data through a wireless network.

[0067] The temperature and humidity sensors are high-precision temperature and humidity sensors based on digital interfaces (such as I2C and SPI), and are arranged in different areas of the workshop, with a quantity of N1. The infrared optical sensors are uncooled infrared sensors, such as the Melexis MLX90614 series, with a detection range of -40 to 125°C; they are arranged near key monitoring equipment, with a quantity of N2. The acoustic sensors are MEMS microphones with a frequency response range of 20Hz to 20kHz and a sensitivity of -42dB or higher; they are arranged near noise sources, with a quantity of N3. The gas sensors are MOS or MOX gas sensors, which can detect VOCs, CO, and NO2; they are arranged near exhaust vents and ventilation openings, with a quantity of N4.

[0068] An adaptive data acquisition strategy is constructed, comprising an event-triggered sampling strategy and an adaptive sampling strategy; the adjustment step size of the sensor sampling time interval is defined as t0, and the initial sampling interval is t; the current sampling interval is set to t. i The current data is d i The previous data was d i-1 .

[0069] The event-triggered sampling strategy includes: when When the sensor is triggered to sample, the sampled data is updated; at the same time, the current sampling interval t is changed. i The value is reduced by t0, but not less than the set minimum sampling interval t. min ;where δ (g) Let g be the preset threshold for the g-th type of sensor, where g ∈ [1, G] and G is the total number of sensor types.

[0070] The adaptive sampling strategy includes: when m consecutive samples satisfy... At that time, the current sampling interval t will be... i The value increases by t0, but does not exceed the set maximum sampling interval t. max .

[0071] consecutive m samples satisfy At that time, the current sampling interval t will be... i The value is reduced by t0, but not less than the set minimum sampling interval t. min Where m is a preset sensitivity parameter.

[0072] Step S1 comprehensively collects multi-dimensional data of the factory environment through sensors, and constructs event-triggered sampling strategies and adaptive sampling strategies. This enables efficient acquisition of factory environmental data, reduces energy consumption, and provides a data foundation for environmental monitoring and anomaly detection.

[0073] S2: Construct a sensor self-organizing network strategy to automatically discover sensor networks and dynamically form networks.

[0074] A self-organizing network protocol is constructed, treating each sensor as a network node. The network node includes a parent node network composed of network coordinators and a child node network composed of sensors.

[0075] Dynamic addresses are assigned to sensors, and a network address pool is preset in the network coordinator, with an address range of [A]. min A max ], where A min and A max These are the minimum and maximum values ​​for a 16-bit network address; when a new sensor joins the network, it sends an address request to the network coordinator, carrying a random number R in the address request. s After the network coordinator receives the address request, it will assign R... s With a random number R stored within itself c Perform an XOR operation to obtain the hash value H. s : in, This represents the XOR operator.

[0076] H s For address pool length L = A max -A min +1 modulo operation yields the address A assigned to the new sensor. s A s =A min +(H s mod L), where mod represents the modulo operation.

[0077] The network coordinator will A s Assign it to the new sensor and R s Storage is used for subsequent address conflict detection.

[0078] The system enables automatic sensor discovery and dynamic networking. After a new sensor is powered on, it first enters a network listening state and scans multiple channels for a predetermined time. If no network beacon frames are received, the new sensor acts as a network coordinator, creating new networks on idle channels and periodically broadcasting beacon frames. If the new sensor receives a network beacon frame, it parses the network parameters in the beacon, joins an existing network, and requests address allocation.

[0079] After a new sensor successfully joins the network, it periodically exchanges link quality information with neighboring sensors; and defines the link quality metric Q. ab , representing the link quality between sensor a and sensor b, with a value range of [0, 1]; Q ab The value of Q is calculated by sensor a based on the received signal strength indication (RSSI) and signal-to-noise ratio (SNR) measurements. ab Below the preset threshold Q thAt that time, sensor a actively disconnects from sensor b.

[0080] Step S2 enables automatic discovery and dynamic networking of sensor networks by constructing a sensor self-organizing network protocol, simplifying network deployment and maintenance, improving network flexibility and scalability, and introducing link quality assessment to adaptively optimize network performance.

[0081] S3: Create a link quality adaptive mechanism, periodically detect wireless link signal indicators, and adaptively adjust the transmit power or switch channels when the link quality drops below the threshold.

[0082] Define link quality assessment metrics, and define Signal Strength Indicator (RSSI): RSSI a This represents the signal strength received by sensor a; the unit is dBm; the signal-to-noise ratio (SNR) is defined as: SNR a The signal-to-noise ratio received by sensor a is expressed in dB; the packet loss rate (PLR) is defined as follows: This represents the packet loss rate from sensor a to network coordinator a0, with a value range of [0, 1].

[0083] Link quality is periodically monitored. Each network coordinator periodically broadcasts link probe packets, which include the node number, transmit power, and timestamp. Upon receiving the link probe packets, the child node sensors of the network coordinator measure the Signal Strength Indicator (RSSI) and Signal-to-Noise Ratio (SNR), and send the measurement results, along with their own node number and timestamp, to the parent network coordinator. After receiving the response from child node sensor a, the parent network coordinator a0 calculates the packet loss rate. Where, N sent N represents the total number of packets sent from parent node network coordinator a0 to child node sensor a. received This represents the number of packets successfully received by child node sensor a.

[0084] Define link quality thresholds, including: Signal Strength Indicator (RSSI) threshold. th Signal-to-noise ratio (SNR) threshold th and packet loss rate threshold PLR th .

[0085] When the signal strength indicator RSSI a <RSSI th When the link signal strength is insufficient, the signal-to-noise ratio (SNR) is also insufficient. a <SNR th When the link signal-to-noise ratio is insufficient, the packet loss rate will be high. If this happens, the packet loss rate of the link will be too high.

[0086] The transmit power between the sensor and the network coordinator is adaptively adjusted. When the parent node network coordinator a0 detects that the link quality with the child node sensor a does not meet the quality threshold condition, the transmit power is increased.

[0087] Define the transmit power adjustment step size ΔP, in dBm; let the initial transmit power be P0; when RSSI is satisfied... a <RSSI th or SNR a <SNR th If the link quality still does not meet the quality threshold condition after increasing the transmit power, continue to increase it until the maximum transmit power P is reached. max .

[0088] When child node sensor a detects that the link quality with parent node network coordinator a0 does not meet the quality threshold condition and has reached the maximum transmit power, the channel is switched.

[0089] Define the candidate channel set C = {c1, c2, ..., c K}, where K is the total number of candidate channels; the child node sensor a randomly selects a candidate channel c from the candidate channels. k Then switch to the alternative channel for communication.

[0090] If the link quality still does not meet the quality threshold after switching channels, random switching will continue until all alternative channels have been switched.

[0091] If none of the alternative channels meet the quality threshold conditions, then the child node sensor a sends a network reorganization or node reassociation request to the parent node network coordinator a0.

[0092] A link quality recovery mechanism is constructed. When the child node sensor a detects that the link quality with the parent node network coordinator a0 has recovered to above the quality threshold, the transmit power is reduced successively to reduce energy consumption. The transmit power is reduced by ΔP each time until the initial transmit power P0 is reached. If the link quality deteriorates again after reducing the transmit power, the adaptive adjustment of transmit power or channel switching procedure is re-executed.

[0093] Step S3 creates a link quality adaptive mechanism. By periodically detecting wireless link signal indicators, it determines whether the link quality has dropped below a threshold, automatically adjusts the transmit power or switches channels, ensures the reliability of data transmission, and improves the robustness of the wireless sensor network.

[0094] S4: Extract features from environmental data collected by sensors, build and train a machine learning model to predict environmental data change trends. Extract features from real-time collected environmental data and use the trained machine learning model to predict environmental data change trends.

[0095] Acquire historical environmental data and define the characteristics of the environmental data collected by the sensors, including: temperature data characteristics, humidity data characteristics, infrared optical data characteristics, acoustic data characteristics, and gas concentration data characteristics.

[0096] The temperature data features include: average temperature Temperature variance and the temperature change rate T'; the humidity data characteristics include: average humidity Humidity variance and humidity change rate H'; the infrared optical data features include: average light intensity Light intensity variance And the rate of change of light intensity I'; the acoustic data characteristics include: average sound pressure level sound pressure level variance and the sound pressure level change rate P'; the gas concentration data characteristics include: average gas concentration gas concentration variance And the rate of change of gas concentration C'.

[0097] The acquired historical environmental data is preprocessed, including data cleaning, outlier handling, and data normalization.

[0098] The sliding window method is used to extract time-series features from historical environmental data; a machine learning model is constructed, and the Long Short-Term Memory (LSTM) network is used to model and predict environmental data.

[0099] The input to the LSTM model is the extracted environmental data time series features X = (x1, x2, ..., x...). E ), where E is the number of time windows; the output of the LSTM model is the predicted environmental data for the next F time windows. in Let f be the predicted environmental data value for the f-th time window, where f∈[1,F];

[0100] The acquired historical environmental feature data is divided into training and test sets. The LSTM model is trained using the training set data, and the predictive performance of the LSTM model is evaluated using the test set data. During training, the mean squared error (MSE) is used as the loss function. Where Y = (y1, y2, ..., y F () represents the actual values ​​of real environmental data.

[0101] The trained LSTM model is used to predict real-time environmental data. The specific steps are as follows: preprocessing and feature extraction are performed on the real-time environmental data to obtain the feature vector x of the current time window. t ; will x t The data is input into an LSTM model to obtain predicted environmental data for the next F time windows. Predicted values The data is compared with a set threshold for abnormal trends in environmental data to determine whether the trend of environmental data changes is abnormal.

[0102] Step S4 extracts the characteristics of environmental data collected by sensors and uses machine learning models to predict the changing trends of environmental data. This can help detect abnormal environmental changes in advance, provide a basis for environmental regulation decisions, and improve the foresight and initiative of factory environmental management.

[0103] S5: Construct a factory environment change trend adjustment strategy to pre-adjust the factory environment when the environmental data change trend predicted by the machine learning model is abnormal.

[0104] Define thresholds for abnormal trends in environmental data, including the temperature change trend threshold ΔT. th Humidity change trend threshold ΔH th Infrared light intensity change trend threshold ΔI th Sound pressure level change trend threshold ΔP th and the threshold value ΔC for the trend of gas concentration change th .

[0105] The trained LSTM model is used to predict environmental data for the next F time windows, resulting in a sequence of predicted values. in These represent the predicted values ​​for temperature, humidity, infrared light intensity, sound pressure level, and gas concentration for the f-th time window, respectively.

[0106] Calculate the trend of the predicted value series within the predicted time window: temperature change trend Humidity change trend Infrared light intensity variation trend Sound pressure level change trend Gas concentration change trend Where Δt is the time interval between adjacent prediction time windows.

[0107] Perform anomaly detection on the predicted trend sequence: If The predicted temperature change trend is abnormal; if The predicted humidity change trend is abnormal; if ΔI th If the predicted trend of infrared light intensity change is abnormal; if The predicted sound pressure level change trend is abnormal; if The predicted trend of gas concentration change is abnormal.

[0108] If any outlier appears in the predicted trend sequence, the factory environment change trend adjustment strategy will be implemented to restore the corresponding environmental data change trend to within the outlier trend threshold.

[0109] When temperature changes abnormally, the air conditioning system is used to adjust the factory's ambient temperature to restore it to normal. When humidity changes abnormally, a dehumidifier or humidifier is activated to adjust the humidity to normal. When infrared light intensity changes abnormally, the infrared light source equipment is inspected and repaired, and the infrared light intensity in the factory environment is adjusted to normal. When sound pressure level changes abnormally, the sound source equipment is inspected and repaired, and the noise level in the factory environment is adjusted to normal. When gas concentration changes abnormally, the ventilation system is activated to adjust the gas concentration in the factory environment to normal.

[0110] After environmental adjustments, the LSTM model continues to predict future environmental data. If the predicted trend remains abnormal and meets alarm conditions, an alarm is issued to factory management, indicating that the abnormal environmental trend cannot be automatically adjusted and requires manual inspection and handling. The alarm information includes the time of the anomaly, the type of anomaly (temperature, humidity, infrared light intensity, sound pressure level, or gas concentration change trend), and the degree of anomaly (the extent to which the predicted trend value exceeds the abnormal trend threshold). The alarm conditions are as follows:

[0111]

[0112] Where I(·) is an indicator function, which takes the value of 1 when the absolute value of the predicted trend exceeds the corresponding abnormal trend threshold, and takes the value of 0 otherwise. This represents the predicted trend of environmental data changes in the f-th time window, where X represents any type of environmental data, and ΔX th α is the corresponding abnormal trend threshold; α is the abnormal proportion threshold.

[0113] Step S5 establishes a factory environment change trend adjustment strategy. Based on the abnormal environmental change trend predicted by the machine learning model, it adjusts environmental parameters in a timely manner through automated adjustment equipment to restore the environment to normal, reduce the impact of abnormal environment on production, and improve the level of factory environmental management.

[0114] Example 2

[0115] Reference Figure 2 This is a second embodiment of the present invention, which provides a factory environment monitoring system integrating a sensor network.

[0116] The system includes: a data acquisition strategy module, an automatic networking module, a network signal conditioning module, an environmental data prediction module, and an environmental conditioning module.

[0117] The data acquisition strategy module collects multi-dimensional data of the factory environment through sensors and constructs an adaptive data acquisition strategy, including an event-triggered sampling strategy and an adaptive sampling strategy.

[0118] The automatic networking module is used to construct a sensor self-organizing network strategy, automatically discover sensor networks, and dynamically form a network.

[0119] The network signal conditioning module is used to create a link quality adaptive mechanism, periodically detect wireless link signal indicators, and adaptively adjust the transmission power or switch channels when the link quality drops below a threshold.

[0120] The environmental data prediction module is used to extract features from environmental data collected by sensors, construct and train a machine learning model to predict the trend of environmental data changes. It extracts features from real-time collected environmental data and uses the trained machine learning model to predict the trend of environmental data changes.

[0121] The environmental adjustment module is used to construct a factory environment change trend adjustment strategy. When the environmental data change trend predicted by the machine learning model is abnormal, the factory environment is pre-adjusted.

[0122] In addition, the parts of the technical solutions provided in the embodiments of the present invention that are consistent with the implementation principles of the corresponding technical solutions in the prior art have not been described in detail, so as to avoid excessive elaboration.

[0123] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A factory environment monitoring method integrating a sensor network, characterized in that, include: Multi-dimensional data of the factory environment is collected through sensors, and an adaptive data acquisition strategy is constructed, including an event-triggered sampling strategy and an adaptive sampling strategy. Build a sensor self-organizing network strategy to automatically discover sensor networks and dynamically form networks; Dynamic addresses are assigned to sensors. The network coordinator presets a network address pool, calculates the network address assigned to each sensor, and performs network address conflict detection on the sensors. Dynamic addresses are assigned to sensors, and a network address pool is preset in the network coordinator, with an address range of [ ],in and These are the minimum and maximum values ​​for a 16-bit network address; when a new sensor joins the network, it sends an address request to the network coordinator, carrying a random number in the address request. After the network coordinator receives the address request, it will With a random number stored within itself Perform an XOR operation to obtain the hash value. : ,in, This represents the XOR operator; Will Address pool length Modulus is taken to obtain the address assigned to the new sensor. : ,in, This represents the modulo operation; The network coordinator will Assigned to the new sensor, and Storage is used for subsequent address conflict detection; Create a link quality adaptive mechanism to periodically detect wireless link signal indicators. When the link quality drops below a threshold, adaptively adjust the transmit power or switch channels. Define link quality assessment metrics and define signal strength indicators. , indicating sensor Received signal strength; define signal-to-noise ratio. , indicating sensor Received signal-to-noise ratio; define packet loss rate , indicating sensor To network coordinator Packet loss rate; Link quality is periodically checked. Each network coordinator periodically broadcasts link probe packets, which include the node number, transmit power, and timestamp. Upon receiving the link probe packets, the child node sensors of the network coordinator measure the Signal Strength Indicator (RSSI) and Signal-to-Noise Ratio (SNR), and send the measurement results, along with their own node number and timestamp, to the parent network coordinator. The parent network coordinator... Received child node sensor After receiving the reply, calculate the packet loss rate. : ;in, The parent node is the network coordinator. Send to child node sensors Total number of packages For child node sensors Number of packets successfully received; Define link quality thresholds, including: signal strength indication threshold. Signal-to-noise ratio threshold and packet loss rate threshold When the signal strength indicator When the signal strength is insufficient, the link signal strength is also insufficient; when the signal-to-noise ratio is low... When the link signal-to-noise ratio is insufficient, the packet loss rate will be high. At that time, the packet loss rate of the link is too high; Adaptive adjustment of transmit power between the sensor and the network coordinator, when the parent node network coordinator... Detected with child node sensor When the link quality does not meet the quality threshold condition, increase the transmission power; Define transmit power adjustment step size Let the initial transmission power be When satisfied or At that time, the transmission power will be increased. If the link quality still does not meet the quality threshold after increasing the transmit power, continue increasing the transmit power until the maximum transmit power is reached. ; When child node sensor Detected network coordinator with parent node When the link quality does not meet the quality threshold condition and the maximum transmit power has been reached, switch channels; Define the set of alternative channels },in Total number of alternative channels; sub-node sensors Randomly select one alternative channel from the alternative channels. Switch to the alternative channel for communication; If the link quality still does not meet the quality threshold condition after switching channels, random switching will continue until all alternative channels have been switched. If all alternative channels fail to meet the quality threshold condition, then the child node sensor... To the parent node network coordinator Send a network reorganization or node reassociation request; Construct a link quality recovery mechanism when the child node sensor Detected network coordinator with parent node When the link quality recovers to above the quality threshold, the transmit power is gradually reduced, with each reduction in transmit power being... until the initial transmit power is reached. If the link quality deteriorates again after reducing the transmit power, the adaptive transmit power adjustment or channel switching procedure will be re-executed. Extract features from environmental data collected by sensors, build and train a machine learning model to predict the trend of environmental data change. Extract features from real-time collected environmental data and use the trained machine learning model to predict the trend of environmental data change. Develop a strategy to adjust the factory environment in advance when the environmental data predicted by the machine learning model shows abnormal trends.

2. The factory environment monitoring method with an integrated sensor network according to claim 1, characterized in that, In factories, sensors include temperature sensors, humidity sensors, infrared optical sensors, acoustic sensors, and gas sensors. An adaptive data acquisition strategy is constructed, comprising an event-triggered sampling strategy and an adaptive sampling strategy; the adjustment step size of the sensor sampling time interval is defined as follows. The initial sampling interval is Let the current sampling interval be The current data is The last data was ; The event-triggered sampling strategy includes: when When the sensor is triggered to sample, the sampling data is updated; at the same time, the current sampling interval is changed. value reduction But not less than the set minimum sampling interval ;in For the first Preset thresholds for various types of sensors , Total number of sensor types; The adaptive sampling strategy includes: when continuous The second sampling satisfies At that time, the current sampling interval will be... Value increases But not greater than the set maximum sampling interval ; continuous The second sampling satisfies At that time, the current sampling interval will be... value reduction But not less than the set minimum sampling interval ;in These are the preset sensitivity parameters.

3. The factory environment monitoring method with integrated sensor network according to claim 2, characterized in that, A self-organizing network protocol is constructed, treating each sensor as a network node. The network node includes a parent node network composed of network coordinators and a child node network composed of sensors. The system enables automatic sensor discovery and dynamic networking. After a new sensor is powered on, it first enters a network listening state and scans multiple channels for a predetermined time. If no network beacon frames are received, the new sensor acts as a network coordinator, creates a new network on an idle channel, and periodically broadcasts beacon frames. If the new sensor receives a network beacon frame, it parses the network parameters in the beacon, joins an existing network, and requests address allocation.

4. The factory environment monitoring method with an integrated sensor network according to claim 3, characterized in that, Acquire historical environmental data and define the characteristics of the environmental data collected by the sensors, including: temperature data characteristics, humidity data characteristics, infrared optical data characteristics, acoustic data characteristics, and gas concentration data characteristics; The temperature data features include: average temperature Temperature variance and the rate of temperature change The humidity data features include: average humidity. Humidity variance and humidity change rate The infrared optical data features include: average light intensity. Light intensity variance and the rate of change of light intensity The acoustic data characteristics include: average sound pressure level. Sound pressure level variance and the rate of change of sound pressure level The gas concentration data features include: average gas concentration. Gas concentration variance and the rate of change of gas concentration '; The sliding window method is used to extract time-series features from historical environmental data, a machine learning model is constructed, and a Long Short-Term Memory (LSTM) network model is used to model and predict environmental data.

5. The factory environment monitoring method with an integrated sensor network according to claim 4, characterized in that, The input to the LSTM model is the extracted time-series features of environmental data. ,in The number of time windows in the historical environmental feature dataset; the output of the LSTM model is the future... Environmental data prediction values ​​for each time window ,in For the first Environmental data prediction values ​​for a time window ; The acquired historical environmental feature data is divided into a training set and a test set. The LSTM model is trained using the training set and evaluated using the test set. During training, mean squared error is used as the loss function. The trained LSTM model is used to predict real-time acquired environmental data. The specific steps are as follows: preprocessing and feature extraction are performed on the real-time acquired environmental data to obtain the current time series windowed feature vector. ,Will Input into the LSTM model to obtain the future Environmental data prediction values ​​for each time window ; predict the value The data is compared with a set threshold for abnormal trends in environmental data to determine whether the trend of environmental data changes is abnormal.

6. The factory environment monitoring method with an integrated sensor network according to claim 5, characterized in that, Define thresholds for abnormal trends in environmental data, including thresholds for temperature change trends. Humidity change trend threshold Infrared light intensity change trend threshold Sound pressure level change trend threshold and the threshold for gas concentration change trends ; Using a trained LSTM model to predict the future Environmental data within a time window are used for prediction to obtain a sequence of predicted values. ,in They represent the first The predicted values ​​for temperature, humidity, infrared light intensity, sound pressure level, and gas concentration for each time window.

7. The factory environment monitoring method with an integrated sensor network according to claim 6, characterized in that, The calculation shows the changing trends of environmental data within a time window, including temperature change trends. Humidity change trend Infrared light intensity variation trend sound pressure level change trend Gas concentration change trend ; Perform anomaly detection on the predicted trend sequence: If If so, the predicted temperature change trend is abnormal; if If so, the predicted humidity change trend is abnormal; if If the predicted trend of infrared light intensity change is abnormal; if If the predicted sound pressure level change trend is abnormal, then... If so, the predicted trend of gas concentration change is abnormal; If any anomaly occurs in the predicted trend sequence, the factory environment change trend adjustment strategy will be implemented to restore the corresponding environmental data change trend to within the abnormal trend threshold. After environmental adjustments, the LSTM model continues to be used to predict future environmental data. If the predicted trend remains abnormal and alarm conditions are met, an alarm is issued. The alarm conditions are as follows: ； in For indicator functions; Indicates the predicted first Environmental data change trends over a time window Refers to any type of environmental data. The corresponding abnormal trend threshold; This is the threshold for the abnormal proportion.

8. A factory environment monitoring system with an integrated sensor network, used to implement the factory environment monitoring method with an integrated sensor network as described in any one of claims 1 to 7, characterized in that, include: The system includes a data acquisition strategy module, an automatic networking module, a network signal conditioning module, an environmental data prediction module, and an environmental conditioning module. The data acquisition strategy module collects multi-dimensional data of the factory environment through sensors and constructs an adaptive data acquisition strategy, including an event-triggered sampling strategy and an adaptive sampling strategy. The automatic networking module is used to construct a sensor self-organizing network strategy, automatically discover sensor networks, and dynamically form a network. The network signal conditioning module is used to create a link quality adaptive mechanism, periodically detect wireless link signal indicators, and adaptively adjust the transmission power or switch channels when the link quality drops below a threshold. The environmental data prediction module is used to extract features of environmental data collected by sensors, construct and train a machine learning model to predict the trend of environmental data changes. It extracts features from real-time collected environmental data and uses the trained machine learning model to predict the trend of environmental data changes. The environmental adjustment module is used to construct a factory environment change trend adjustment strategy. When the environmental data change trend predicted by the machine learning model is abnormal, the factory environment is pre-adjusted.

Images